Yi Qian
Senior Software Engineer at AMD
Frisco, Texas, United States
Summary
🤩
Rockstar🎓
Top SchoolYi Qian is a Senior Software Engineer and compiler specialist with over two decades of experience in high-level optimizations, compiler design, and SSA-based transformations, currently driving compiler work at AMD. He combines deep research credentials (Ph.D. in Computer Science) with practical product delivery, having spent 18 years at Texas Instruments optimizing DSP code, defining instruction sets, and implementing vectorization and loop transformations. At AMD he has contributed to machine-specific support in the Triton compiler—adding MI300 integration tests, buffer cache modifiers, and gfx950 MFMA support—demonstrating hands-on expertise with GPU microarchitectures. Known for methodical problem solving and rapid technology adoption, he bridges low-level code generation and high-level optimization strategy. Colleagues rely on his detail-oriented approach to debug complex passes like RewriteTensorPointer and to add numeric conversions for emerging FP formats. Based in Frisco, Texas, he pairs strong communication skills with a track record of shipping robust compiler features for real hardware.
5 years of coding experience
18 years of employment as a software developer
Doctor of Philosophy (Ph.D.), Computer Science, Doctor of Philosophy (Ph.D.), Computer Science at Michigan Technological University
English, Chinese
Github Skills (8)
triton10
c-language10
backend10
cprogramming-language10
back-end-development10
llvm10
mlr9
hip9
Programming languages (6)
Github contributions (5)
triton-lang/triton
May 2024 - Mar 2025
Development repository for the Triton language and compiler
Role in this project:
Backend Developer
Backend DeveloperContributions:15 reviews, 30 PRs, 24 comments in 9 months
Contributions summary:Yi's primary contributions focused on integrating integration tests for AMD MI300 within the Triton language and compiler repository. This involved updating CI scripts and addressing test failures specific to the MI300 architecture, demonstrating a strong understanding of the target platform. Furthermore, the user addressed issues in the RewriteTensorPointer pass and added cache modifiers to BufferLoad and BufferStore operations. The user also added fp32 to OCP fp8/bf8 conversions for the MI350 and added support for gfx950 double rate mfma operations.
compilerprogramming-languagecode-generationtriton
yiqian1/triton
May 2024 - Apr 2025
Development repository for the Triton language and compiler
Contributions:122 pushes, 28 branches in 10 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.